Trustworthy AI Course: Technical Interpretation of ESSAI'26 Summer School

Professor Indrė Žliobaitė from the University of Helsinki will teach a Trustworthy AI course at the 2026 European Summer School on Artificial Intelligence (ESSAI'26), covering core topics such as interpretability, fairness, and causal machine learning. The course is aimed at interdisciplinary early-career researchers, aiming to address ethical challenges in AI decision-making applications from a technical perspective, with both theoretical depth and industry practical value.

With the widespread application of AI in areas such as credit approval, public policy, and medical diagnosis, Trustworthy AI (involving fairness, transparency, and accountability) has become a focus of academia and industry. This course is aimed at interdisciplinary audiences interested in the social impact of AI (including researchers in social sciences, humanities, law, etc.), with no formal mathematical or computational background required, and aims to promote interdisciplinary dialogue and collaboration.

Professor Indrė Žliobaitė is a professor in the Department of Computer Science at the University of Helsinki, with over 5 years of teaching experience in trustworthy machine learning. Her master's course has been offered for many consecutive years and is widely praised. Her research interests include fairness-aware machine learning, explainable AI, and causal inference, with numerous publications in top academic journals and conferences.

The course consists of four lectures: 1. Introduction to Trustworthy AI (definition of interpretability, black-box vs. interpretable models, legal requirements); 2. Foundations of Fairness (definitions, historical cases such as redlining, calculation of fairness metrics); 3. Advanced Fairness (contradictions in measurement methods, pre-processing/ in-processing/ post-processing fairness algorithms); 4. Causal Inference and Scientific AI (definition of causality, Do-calculus, scientific application cases such as vegetation prediction for climate change).

Industry insights from the course content: 1. Regulatory compliance (helping enterprises comply with frameworks such as the EU AI Act); 2. Risk management (identifying reputation and legal risks of AI systems); 3. Product design (integrating Trustworthy AI principles to build user-trusted products); 4. Cross-departmental collaboration (promoting effective communication between technical and legal/ethical teams).

The ESSAI'26 Trustworthy AI course represents an important direction in AI education—balancing technical competence and social responsibility cultivation. As AI's role in social decision-making increases, understanding and implementing Trustworthy AI principles will become a core competency for AI practitioners, and this course provides a valuable learning opportunity for relevant researchers.